metadata
dataset_info:
features:
- name: _id
dtype: string
- name: url
dtype: string
- name: title
dtype: string
- name: text
dtype: string
- name: emb
sequence: float32
- name: tokens_count
dtype: int64
splits:
- name: train
num_bytes: 25921932467
num_examples: 5637930
download_size: 12758634117
dataset_size: 25921932467
configs:
- config_name: default
data_files:
- split: train
path: data/train-*